AI-Driven Network Ops| Pipeline Magazine

From: Fabio Gavioli,, Claudio Pasera

Complexity is expensive, but it is a reality in network performance and operations today. To successfully perform 5G services with high performance, the mobile network requires an exceptional transport infrastructure capable of meeting the needs of various services, from capacity and latency to synchronization and more. The network must operate with the highest reliability.

However, even this is not enough. When the unexpected happens, a quick and efficient operational response makes the difference between smooth network operation and interruptions that interrupt traffic. This article examines the composition of the 5G backhaul, the complexity of the network, and the transforming potential of artificial intelligence in networks.

Today, according to a study by GSMA and IT WILL HELP, microwave radio technology accounts for over 65 percent of global global mobile communications. The use of microwave radio as the preferred feedback technology is expected to remain in the foreseeable future (projected to be over 60% in 2027).

On average, a European mobile operator has more than 25,000 microwave radios across the country, generating more than one million event recordings per day, plus several performance measurement points that generate group data that can easily reach one to two billion recordings per year.

The analysis of such a huge amount of data is a titanic task – and that is why it is usually done by operators in a reactive way. Once an event occurs, the data is checked for troubleshooting to correct the interruption and to restore network operations.

Thanks to the use of applications authorized with AI and ML, raw big data can be analyzed through an artificial neural network (ANN), moving to an automated and sound approach to predict network behavior, recognize known network patterns and intervene before damage occurs. be. In the event of an unexpected event, a violent reaction can be introduced, avoiding critical escalations and maximizing network uptime.

For some time now, there has been an evolution towards a new network management model – SDN or software-defined networks. SDN is a software module called a domain controller that acts as a network operating system. This software module collects all the data from the network elements and abstracts the interfaces to manage them to higher levels of management, which allows applications to easily interact with the network.

In most cases, the operator uses two or three providers to build the transmission network, which leads to the creation of its own solution, based on many providers, based on OSS. But it’s worth the price.

In the pre-SDN world, each provider provides microwave radio communications, as well as its own OSS solution for managing, monitoring and collecting equipment performance data, based on its own protocol implementation.